Model S 2017 75D wont sleep properly 1hr sleep, 15min awake repeat...

[vanhtu1987]

My personal concern is more about long term damage to my HV battery from the constant cycling

The degradation is at its worst at high temperature and high rate of discharge. With this type of drain the effect would be negligible.

I don't have a directly relevant study to the effect of low-C discharge on battery degradation, but we could gleam some information from https://core.ac.uk/download/pdf/323272674.pdf (page 42), which shows the degradations at 1/2/3C. Standby power consumption at 50W is otherwise only a ~0.001C discharge rate.

The energy consumption over a year is still significant though, but there's not much we could do about this. It seems to have been improved in recent Model 3/Y models. I've also seen this issue reported for Model S with MCU2 updates (which mine is).

 

[fetzu]

The vampire drain averages 30W (16-20W for a more modern Model 3/Y as they can sleep for longer), over a year it'd cost ~20 quid. This doesn't seem worth concerning ourselves over.

I know that’s what the theory says. My Model S 75D, with/after the MCU 2 upgrade, was loosing up to 6% of battery from idling a day. That’s more than 4 kWh a day (almost 170 W). It is not really the money for the lost energy I am concerned about, but more the fact that I can’t expect to leave my car idle for two weeks and being able to drive off at the end, or the constant cycles it is putting on my battery pack.

 

[vanhtu1987]

I know that’s what the theory says. My Model S 75D, with/after the MCU 2 upgrade, was loosing up to 6% of battery from idling a day. That’s more than 4 kWh a day (almost 170 W). It is not really the money for the lost energy I am concerned about, but more the fact that I can’t expect to leave my car idle for two weeks and being able to drive off at the end, or the constant cycles it is putting on my battery pack.

It shouldn't have consumed that much energy on standby unless you have things like Sentry Mode/Cabin Overheat Protection turned on. Other cars don't have these features so a fair comparison could be made only without them enabled.

As an example, here is my standby graph from yesterday. It lost about 1% which seems to be on par with what other people have been saying (who have been happy with their cars' standby energy consumption).

 

[Droschke]

My car's info can be found in my signature and has been upgraded to MCU2 and a brand new HV battery. The car does not sleep, has connection activity 24 hours and loses 2% (5 miles) of capacity in 24 hours, while parked, unplugged and unused.

 

[Droschke]

My car's info can be found in my signature and has been upgraded to MCU2 and a brand new HV battery. The car does not sleep, has connection activity 24 hours and loses 2% (5 miles) of capacity in 24 hours, while parked, unplugged and unused.

View attachment 837037

Forgot to mention, I do not use any 3rd party app or portal. This behavior started right after MCU2 upgrade. Per service center, my 12v battery is healthy (I had them check its health during the MCU2 upgrade install).

 

[vanhtu1987]

My car's info can be found in my signature and has been upgraded to MCU2 and a brand new HV battery. The car does not sleep, has connection activity 24 hours and loses 2% (5 miles) of capacity in 24 hours, while parked, unplugged and unused.

View attachment 837037

That seems typical for Model S with MCU 2 upgrade. The graph seems to show activities on an hourly basis, which does not necessarily mean that the car doesn't sleep, just that it wakes up at least once at hour.

2% per day for your battery seems a little high (75W standby consumption), but the percentage point is too coarse to say one way or the other. If the car is awake then the power consumption would be more like 150W.

 

[Droschke]

That seems typical for Model S with MCU 2 upgrade. The graph seems to show activities on an hourly basis, which does not necessarily mean that the car doesn't sleep, just that it wakes up at least once at hour.

2% per day for your battery seems a little high (75W standby consumption), but the percentage point is too coarse to say one way or the other. If the car is awake then the power consumption would be more like 150W.

Thanks. That ~2% is ~5 miles per 24 hours. That's with a brand new 90 kWh battery with 80 kWh usable.

How would you arrive at "75W standby consumption" and "150w"?

 

[vanhtu1987]

Thanks. That ~2% is ~5 miles per 24 hours. That's with a brand new 90 kWh battery with 80 kWh usable.

How would you arrive at "75W standby consumption" and "150w"?

I was going with a total capacity of 90kWh, but for 80kWh the power consumption is 0.02 * 80000 / 24 = 66.67W which could add up over the course of a year (584kWh or ~2000 miles..).

If my car is awake for an extended period then it's always due to Sentry Mode keeping it awake. Looking at a sample of the data:

All the standby periods were when the consumption was lower than 100W, on average they're about 35W so I'll lose 35 * 24 / 72600 = 1.15% of battery capacity per day, which is right inline with other people's report (mine is a S75D with 72.6kWh usable, in reality it's got about 67.5kWh usable now so the actual consumption is a little less).

With Sentry Mode, my car consumed ~180W on average. I'd expect that Sentry Mode consumes more power, particularly if lots of events are logged. The 150W figure is just a guesstimate, we'll need to keep it awake by keeping the Tesla app open, which I don't have the data for.

Edit: Perhaps we could scale the data based on the sleep/awake ratio. Mine appears to sleep for 57 minutes and wakeup for 15 minutes so the power consumption is 35 * ((57 + 15) / 15) = 168W, or 5.3% of battery per day (72.6kWh) if the car is awake all the time.

 

[Droschke]

I was going with a total capacity of 90kWh, but for 80kWh the power consumption is 0.02 * 80000 / 24 = 66.67W which could add up over the course of a year (584kWh or ~2000 miles..).

If my car is awake for an extended period then it's always due to Sentry Mode keeping it awake. Looking at a sample of the data:
View attachment 837204

All the standby periods were when the consumption was lower than 100W, on average they're about 35W so I'll lose 35 * 24 / 72600 = 1.15% of battery capacity per day, which is right inline with other people's report (mine is a S75D with 72.6kWh usable, in reality it's got about 67.5kWh usable now so the actual consumption is a little less).

With Sentry Mode, my car consumed ~180W on average. I'd expect that Sentry Mode consumes more power, particularly if lots of events are logged. The 150W figure is just a guesstimate, we'll need to keep it awake by keeping the Tesla app open, which I don't have the data for.

Edit: Perhaps we could scale the data based on the sleep/awake ratio. Mine appears to sleep for 57 minutes and wakeup for 15 minutes so the power consumption is 35 * ((57 + 15) / 15) = 168W, or 5.3% of battery per day (72.6kWh) if the car is awake all the time.

Thanks again. The yearly ~584kWh vampire drain is high. The other factor (disadvantage) seems to be the inability of my car's MCU2 to take advantage of the build-in power supply in the new 90kWh pack due to lack of required wiring (for cars built pre-summer 2015) and that the MCU2 is powered not by the 12v batery but by the HV pack, resulting in higher vampire drain. My concern is the hourly closing/opening of the contactor leading to its premature wear and tear.

 

[sb1089]

Never the less, as soon as I change my password which deactivates any connected API and only login to the Tesla app on my iPhone, the hourly clicking sound sound is gone and the Tesla iOS widget doesn’t update for multiple hours.
As soon as I connect just 1 app (I tried Teslamate first, and after another password change I tried the Home Assistant integration), the same hourly clicking sound is back and the widget also updates itself every hour and a few minutes.

I do really think it has got to do to something with external connected apps (in my case). Even though polling shouldn’t wake my car.

 

[GtiMart]

Ah, yes. External apps that are not well behaved will definitely wake your car. Most api calls wake the car. I know TeslaFi, when configured properly, does not wake my car. But Ive heard people that have trouble.

If you want to use teslamate, I suspect there is configuration on the polling. You might want to look at your config

 

[sb1089]

I tried to diagnose it a bit more.
I disabled Teslamate and the hourly HV clicking stopped. Also the Tesla iOS widget stopped updating the state every 90 minutes (which it did when Teslamate was connected).
Now as soon as I start up Teslamate, the same issue comes back. Hourly clicking noise, the iOS widget updating every ±90 minutes and Teslamate keeps saying my car comes back online after 57 minutes.

Now,
I disabled Teslamate again and tried to monitor what the API did (by using Postman).
So if I do a get on api/1/vehicles, I get a list of vehicles (which is only my Model S). The response also states if my car is online or offline.
Now every hour, my vehicle seems to come into the state 'online'

{
"response": [
{
"id": XXXXX,
"vehicle_id": XXXX,
"vin": "XXXXX",
"display_name": "Charles",
"option_codes": "XXXXX",
"color": null,
"access_type": "OWNER",
"tokens": [
"XXXXX",
"XXXXX"
],
"state": "offline",
"in_service": false,
"id_s": "---",
"calendar_enabled": true,
"api_version": 42,
"backseat_token": null,
"backseat_token_updated_at": null
}
],
"count": 1
}

So the state is 'online' for a maximum of 2 minutes and is offline again.
But.. the car doesn't do the HV clicking and my iOS widget doesn't update as it seems not to retrieve any information from my car (for 'really' waking up the car).

So the problem is, that any connected application, just like Teslamate, sees the online state and immediately starts polling the car for a bit.
But this will actually cycle the car to be online and stream data, which triggers the HV battery again and so on.

Now I don't know what really can be a fix, other than that Teslamate maybe ignore the online state within <120 seconds and if possible only for the cars <2018 with the MCU upgrade, because I think it only affects those cars.

Other note, if I manually trigger the API to wake the car, you hear the click and the car stays in the state online for more than 120 seconds (about 3 minutes at the time I tested it).

Seems like some kind of bug/result in de Tesla API, but if the connected app can ignore the online state for let's say 120 sec, I have the idea that it will behave with less vampire drain.
I also posted it to the Teslamate github. Car constantly cycles sleep and wake. · Issue #2556 · adriankumpf/teslamate

 

[fetzu]

I talked to a friendly ranger at a supercharger and he told me the unusual pattern (offline/online hourly cycles) was voluntarily introduced by Tesla to fix an issue with the new Swiss telecom provider (Sunrise). Apparently they have to do this to maintain connectivity. Take this with a grain of salt (since my own car “wakes up” offline about 35% of the time).

 

[Droschke]

Take this with a grain of salt

Done

 

[ZzRicezZ]

Installed Teslamate and also notice the same sleep pattern as most for the Model S 75D (MCU2 upgraded). But my M3P is unaffected by Teslamate? So it gotta be something the 75D is doing.

 

[Droschke]

Installed Teslamate and also notice the same sleep pattern as most for the Model S 75D (MCU2 upgraded). But my M3P is unaffected by Teslamate? So it gotta be something the 75D is doing.

View attachment 860210

That's how my S85 has been behaving since I upgraded to MCU2. See Model S 2017 75D wont sleep properly 1hr sleep, 15min awake repeat...

 

[Maurice1]

Installed Teslamate and also notice the same sleep pattern as most for the Model S 75D (MCU2 upgraded). But my M3P is unaffected by Teslamate? So it gotta be something the 75D is doing.

View attachment 860210

Is een vrij normaal patroon na MCU2 upgrade. Dat heb ik ook na de upgrade. In den beginne 3% vampire drain per dag (i.p.v. 1%). De auto heeft tijdens corona 2 weken bij Tesla gestaan voor monitoring. Tesla heeft enkele maanden later een upgrade uitgebracht waardoor alle auto's met een MCU2 upgrade een verminderde vampire drain zouden moeten ervaren. De mijn in elk geval zeker. Nu is het meestal 1% en soms nog eens 2 of 3%.

 

[Droschke]

Tesla released an upgrade several months later that should allow all cars with an MCU2 upgrade to experience a reduced vampire drain. Mine for sure. Now it is usually 1% and sometimes another 2 or 3%.

Which update is that?

 

[Maurice1]

De MCU2 was in okt 21 geïnstalleerd. Pas in April 22, ging de vampire drain naar betere waarden, dus 1 van de volgende updates:
MCU2 was installed in Oktober ' 21. In April 22 the vampire drain reduced to more normal values, so one of the following updates:

 

[Droschke]

De MCU2 was in okt 21 geïnstalleerd. Pas in April 22, ging de vampire drain naar betere waarden, dus 1 van de volgende updates:
MCU2 was installed in Oktober ' 21. In April 22 the vampire drain reduced to more normal values, so one of the following updates:

View attachment 860276

1) Let's stick with English please. You can do it.
2) I've a newer software version than what you are showing (see my signature below) and my vampire drain is much more than when I was on MCU1 (as I've reported above).